Experiments and Inductions. 199 



roots on different surfaces, there to form the interior terminals 

 enclosing a space in which no electric lines exist. I have found 

 this confirmed by experiments made with circular plates 5 inches 

 in diameter stuck on cylinders of lac (fig. 34). Narrow slips of 

 gold-leaf about 1^ inch long were stuck on at three or four 

 points in a line passing through the centre, so that when the 

 plate was charged they rose on end, and would stand thus for an 

 hour (d } q). When the twin-plate p, equally charged, was brought 

 down parallel to q, there was a certain distance (about 3 inches) 

 at which they dropped suddenly dead as it were; and looking 

 through a lens at them when the distance was reduced to about 

 2 inches, there was not the slightest appearance of electric ex- 

 citement at the edges of the leaves when p was tilted a little so as 

 to disturb the general equilibrium of the system of lines. 



It appears that in such cases, when bodies similarly electrified 

 are made to approach, we are compelled to admit the existence 

 of a series of inner terminal lines acting laterally outwards and 

 sideways, and sustained by their contractile tension. (See 60.) 



60. When insulated conductors charged with different electri- 

 cities approach each other, the charge becomes more engaged 

 and less free. The mechanical integrals of the lines diminish, 

 and electric tension is converted into work. When forced sepa- 

 rate it is vice versa. 



Let Q and R (fig. 36) represent two equal spheres segmented 

 at q and r so that there should be equal and parallel circular 

 planes opposing each other at those points. Suppose these 

 spheres to be equally charged with the opposite electricities, and 

 to move backwards and forwards in the line Q R — an impervious 

 dielectric medium, such as a mica plate, intervening between the 

 circles q, r to prevent discharge when close together. 



When Q and R are at a distance, almost all the electricity upon 

 them is free and distributed over their spherical surface (fig. 36) ; 

 when they are close together, almost all the electricity is engaged 

 (fig. 37), and to be found in the opposite circles q } r separated 

 by the thin impervious non-conducting film. While approaching, 

 they are impelled by the prepondering force of the electricities 

 (or root pulling) on the near sides, and work or dynamical force 

 is engendered. The electric lines collapse, and their integral 

 force is expended by being converted into motion. They are to a 

 certain degree discharged ; but complete discharge has not taken 

 place, for the roots are still polarized, and by applying force the 

 spheres may be withdrawn from each other and work reconverted 

 into electric lines. 



Let us consider the equilibrium of the lines in an intermediate 

 position such as fig. 35 : what is the force that packs close 

 together the lines on the near surfaces ? The condition of equi- 



